Killing of tumor cells with cytotoxic drugs is a complex multistep process, which is determined to a large extent by functions of normal cellular genes. Decreased function or expression of some of these genes would lead to increased drug resistance; genetic changes of this type would be manifested as recessive. Only a few recessive mechanisms of drug resistance have been identified; a decrease in the activity of topoisomerase II (topo II) in cells cross-resistant to different topo II poisons provides the best known example of recessive drug resistance. Cloning of novel recessive genes and direct documentation of their function have been hindered by the inability to select for such genes in gene transfer assays. In contrast to the genes themselves, genetic elements derived from recessive drug resistance genes and capable of suppressing their function would behave as dominant selectable markers in gene transfer protocols. Such genetic suppressor elements (GSEs) can therefore serve as tools for studying recessive genetic mechanisms of drug resistance and cloning novel recessive genes. The present proposal is based on a new strategy for selection of efficient GSEs encoding antisense RNA or truncated proteins from a library of randomly fragmented cDNA of the target gene. This strategy was tested out in a bacterial model system and then used in mammalian cells to generate a set of GSEs for the human topo II gene. These GSEs were selected by their ability to confer resistance to etoposide, a topo II poison. The structure and the mechanisms of inhibitory action by these GSEs will now be characterized. The same approach will be utilized to isolate and characterize GSEs derived from topoisomerase I (topo I) cDNA and capable of conferring resistance to camptothecin, a topo I poison. Topo II- and topo I-derived GSEs will be used to study the effects of topoisomerase suppression on cellular resistance to different drugs, cell growth and genomic stability. The principle of random fragment selection will also be expanded to isolate GSEs from presently unknown genes associated with drug resistance. This will be done by generating a random fragment library in a retroviral vector, starting from a normalized (uniform-abundance) cDNA population of human HeLa cells. The usefulness of this conferring etoposide resistance; some of the resulting clones should represent topo II-derived GSEs. The same library will then be used to isolate GSEs conferring resistance to other chemotherapeutic drugs (methotrexate, cisplatin, activated cyclophosphamide and doxorubicin) and to identify the genes corresponding to such GSEs. This approach should help to reveal presently unknown mechanisms for cell killing and drug resistance and may indicate promising targets for the development of new anti-cancer drugs.